Why can't I snap to two end points of an angular line?

[cadkiller]

Group;

I have a site plan with angular property lines and I want to dimension the lines with an aligned dimension; but can't seem to snap to two end points of a line. I can only snap to an end point after my first snap selection; but can't do two end snaps. I have to create a perpendicular line on one end to achieve this.

Why in the world can't I do this?

It would be nice if I could just select the line I want to dimension (similar to Autocad) or even be able to select several lines (similar to Autocad's quick select command) and have them dimensioned automatically.

[Melarch]

When dimensioning to angular line relationships you have to tab to cycle through the dimension location points for the ends of lines.

Mel Persin, AIA

[cadkiller]

Okay I was finally able to get it to work, when using the TAB key to snap to the first end point. For some strange reason the SE (snap end) shortcut doesn't work when snappng to an end point of an object as your first selection. This is very fustrating when something only works at certain times.

[irwin]

Okay I was finally able to get it to work, when using the TAB key to snap to the first end point. For some strange reason the SE (snap end) shortcut doesn't work when snappng to an end point of an object as your first selection. This is very fustrating when something only works at certain times.
The reason this doesn't work is that there are two distinct operations in Revit, selecting and snapping.

Snapping is used when, for example, you are sketching a line. Snapping applies in cases where you could click anywhere on the screen, but geometry is being used to digitize a precise point. The result doesn't especially depend on the geometry that you snap to, though in some cases it might infer certain relationships (such as joining). Snapping also has no associated depth.

On the other hand, selection is an operation that requires there be something for you to select. The thing that you are selecting is an integral part of the operation. In Revit, dimensions are not pieces of dumb geometry that you are drawing on the screen. Rather, they are annotations that reference existing geometry and update when that geometry updates. So, creating a dimension requires selection, rather than snapping. That's why SE doesn't work.

Why is this distinction necessary? In a non-parametric system it might not be necessary. But consider this example in Revit. Suppose that there are two lines that appear to overlap in plan, but are at different heights. If you are sketching a line (on a specified workplane) you can snap to the overlapping lines without caring which of the two you are snapping to -- the geometric result is the same (and in fact Revit will treat them as a single snap rather than two to decrease the number of cases to cycle through). On the other hand, if you are dimensioning to them the software must distinguish which of the two you selected. Otherwise if one of the two lines moves the dimension won't know whether to move with it or not.

Within the selection operation, SE doesn't mean anything, because SE deals with snapping not selection. It might be possible to implement similar keyboard control in selection, where typing SE would mean "filter out everything except endpoints". But, I'm just explaining why it is the way it is now.

[PaulB]

Irwin,

As always, you provide so much information explaining the reasoning behind the way Revit works. I think we are so used to working with "dumb" entities that we expect to be able to do exactly what we have for the last 15 yrs or so using other software when, in fact, if we were to do the same with Revit we would be going backwards.

Your contribution to this forum and to Revit in general is a breathe of fresh air in an age where you usually buy the software and are stuck with very little information or help on the reasoning behind the software.

I can see from this explanation exactly why Revit is so for ahead of the pack.

[jairo.medina]

Aligned dimension for non parallel walls in Revit 8.1:

FIRST PLACE DIMENSION
Dimension
Aligned
Prefer: Wall Faces
Pick : Entire Walls
Options: Openings (Yes), Width (Yes), Intersecting Walls (Yes)
Select a wall
Select blank space to place dimension

EDIT DIMENSION
Select modify
Select dimension
Edit withness lines
Pick end pints of wall
Pick empty space to finish

Repeat with non parallel walls

[steve922542]

But these are not the dimensions that a contractor would use to lay-out these walls. How do you dimension from the face of one of the walls to the inside of one of the corners? Or from the outside corner to an inside corner? I can't seem to produce those type of dimensions.

-Steve Layne

[Mike Sealander]

I do a lot of non-orthogonal geometry, and now use a lot of detail lines to provide orthogonal snaps for dimensions. It's a little frustrating, but it works, and the drawings look nice, because there is a good centerline, for instance, that helps tie the dimension to the point being referenced.

[steve922542]

Ok, so you use the detail line to represent a laser or control line in the field? And then dimension from the control line to the points?

-Steve Layne